Yes, the three IRF fets should be matched. Do you already have the Lovoltech JFETs? If not I would recommend ordering a kit from tech-diy as the kit also includes matched fets.
Last edited:
F3 Build and caps
Hi Thorsten, thanks for pointing those posts out. Believe it or not I have read through the entire thread more than once. It's a addictive reading and there are some very talented amp builders out there.
Your points are well noted on a "fridge size" bank of capacitors. One thing I do know is that the circuit is 80% of the sound so if you don't get that right then you are wasting money on the remaining 20% i.e parts
I am continually impressed with this F5 circuit on the Tad horns. I'm fortunate to own a pair of custom designed SET 45 monoblocks 2 gain stage with very high quality parts which is the best amp I've heard to date. I'm a great fan of SET amps and have owned a variety of 211s, 845 300b, 2A3 and 45 type amps among many other different amps over the years.
The fact that this little F5 amp can get as close sounding to my 45s is quite remarkable.
In its current state I'd say it sounds more like a typical sounding 845 amp. In fact it reminds me very much of the New Audio Frontiers 845 I had a few years ago. Much more power than the SET45 but did not have the ultimate refinement or texture.
So I'm curious to hear the F3 now compared to my SET 45s to see how close a single gain stage full optimized circuit can get. From what I have read it should solve the texture issue with the F5. I don't care about bass because it is being used at 110 db 8 ohm speakers 650 hertz up.
So on the discussion of caps it sounds like nobody has heard the F3 with anything but electrolytics in the signal path given the size of the values. One thing I do know is the coupling caps have quite a big impact on sound. I've rolled many over the years and rolled some mundorfs and audio notes and various others. I also tried a fully charged blackgate as well which frankly was dreadful compared to the others. Now I'm only talking about a single coupling 0.47uf cap and we could hear how poorly the blackgate performed. So that is my concern with the 15k in electrolytics. On solid state amps it may have lesser impact.
In any case I've managed to track down an F3 and hope to audition it in the next couple of weeks. I will then compare that to the F5 then make a decision which one of these amps to build as monoblocks. I may have an impedance issue with my Bent TAP-X preamp with the F3 so will be checking on that carefully.
Cheers
DJ
Hi Thorsten, thanks for pointing those posts out. Believe it or not I have read through the entire thread more than once. It's a addictive reading and there are some very talented amp builders out there.
Your points are well noted on a "fridge size" bank of capacitors. One thing I do know is that the circuit is 80% of the sound so if you don't get that right then you are wasting money on the remaining 20% i.e parts
I am continually impressed with this F5 circuit on the Tad horns. I'm fortunate to own a pair of custom designed SET 45 monoblocks 2 gain stage with very high quality parts which is the best amp I've heard to date. I'm a great fan of SET amps and have owned a variety of 211s, 845 300b, 2A3 and 45 type amps among many other different amps over the years.
The fact that this little F5 amp can get as close sounding to my 45s is quite remarkable.
In its current state I'd say it sounds more like a typical sounding 845 amp. In fact it reminds me very much of the New Audio Frontiers 845 I had a few years ago. Much more power than the SET45 but did not have the ultimate refinement or texture.
So I'm curious to hear the F3 now compared to my SET 45s to see how close a single gain stage full optimized circuit can get. From what I have read it should solve the texture issue with the F5. I don't care about bass because it is being used at 110 db 8 ohm speakers 650 hertz up.
So on the discussion of caps it sounds like nobody has heard the F3 with anything but electrolytics in the signal path given the size of the values. One thing I do know is the coupling caps have quite a big impact on sound. I've rolled many over the years and rolled some mundorfs and audio notes and various others. I also tried a fully charged blackgate as well which frankly was dreadful compared to the others. Now I'm only talking about a single coupling 0.47uf cap and we could hear how poorly the blackgate performed. So that is my concern with the 15k in electrolytics. On solid state amps it may have lesser impact.
In any case I've managed to track down an F3 and hope to audition it in the next couple of weeks. I will then compare that to the F5 then make a decision which one of these amps to build as monoblocks. I may have an impedance issue with my Bent TAP-X preamp with the F3 so will be checking on that carefully.
Cheers
DJ
No doubt, some people will be a bit puzzled with this, but ...
I found that the F3 was a fairly "slow" amp in comparison to the F5, particularly noticeable in horns, very light cone drivers and also headphones!!
So I systematically changed the capacitors to see about changing this aspect of it without losing any of it's magic -It goes a bit like this ...
If you replace the 1uF film cap on the o/p (C8?) to 200uf film (in // to the 10,000uF electro C1) you will notice quite a difference in the sound - it seems to get "faster" then, but it also changes the "balance" of the sound, so you have to go to your Cmultiplier to "fatten" it up a bit - this is where the Silmics or the Nichicon FineGolds come into their own - if not enough, you may have to change the power supply caps to something like the Mundorf's, Jensen 4Ps, Sikorels, etc that "seem to" (sorry!) produce a bit more "weight" in the bass.
However, as you're only driving the horns above 600Hz, don't need to worry about any of this bass balance stuff and can go straight to something like the BHC T-Network caps (or maybe the Slit foils) for C1 plus maybe something as simple as 50uF of film in // to that - maybe the Russian K73's caps will do okay - for some reason, Auri's seem to be a bit "rough" here as they don't seem to like sharing at all - might be because of reasonances - Sonicaps were okay, Mundorf's okay, K73-16s okay, etc
Another point of contention regarding "speeding the amp up" was to replace the 220uF electro at position C3 with a 40uF film cap - surprising thing is that the Auri's are fine here, altho I prefer the big red Jantzens - also, once again, the "balance" of the amp's sound changes again and it takes a bit more 'stuffing about' to set it up the way you want it to sound -
Just to raise a few eyebrows, C2 ended up as a Rifa peg124 (those axial caps) with initially, an R-C snubber across it (0.1 + 22uF K73) and then just one of those Epcos 20uF blue block caps, forget the number) and C4 also became a 470uF Rifa peg124.
Another curious thing was to replace the power resistors with the Manganin PBH 3/10W units from Issabellenhutte in Germany - not exactly the cheapest things but a much softer, more detailed sound than, for example the Caddocks or the Isotan power resistors.
I had quite a bit of fun with this remarkable amp - A bit of an adventure if you want something a bit different to the standard F3.
... my 2 cents
I found that the F3 was a fairly "slow" amp in comparison to the F5, particularly noticeable in horns, very light cone drivers and also headphones!!
So I systematically changed the capacitors to see about changing this aspect of it without losing any of it's magic -It goes a bit like this ...
If you replace the 1uF film cap on the o/p (C8?) to 200uf film (in // to the 10,000uF electro C1) you will notice quite a difference in the sound - it seems to get "faster" then, but it also changes the "balance" of the sound, so you have to go to your Cmultiplier to "fatten" it up a bit - this is where the Silmics or the Nichicon FineGolds come into their own - if not enough, you may have to change the power supply caps to something like the Mundorf's, Jensen 4Ps, Sikorels, etc that "seem to" (sorry!) produce a bit more "weight" in the bass.
However, as you're only driving the horns above 600Hz, don't need to worry about any of this bass balance stuff and can go straight to something like the BHC T-Network caps (or maybe the Slit foils) for C1 plus maybe something as simple as 50uF of film in // to that - maybe the Russian K73's caps will do okay - for some reason, Auri's seem to be a bit "rough" here as they don't seem to like sharing at all - might be because of reasonances - Sonicaps were okay, Mundorf's okay, K73-16s okay, etc
Another point of contention regarding "speeding the amp up" was to replace the 220uF electro at position C3 with a 40uF film cap - surprising thing is that the Auri's are fine here, altho I prefer the big red Jantzens - also, once again, the "balance" of the amp's sound changes again and it takes a bit more 'stuffing about' to set it up the way you want it to sound -
Just to raise a few eyebrows, C2 ended up as a Rifa peg124 (those axial caps) with initially, an R-C snubber across it (0.1 + 22uF K73) and then just one of those Epcos 20uF blue block caps, forget the number) and C4 also became a 470uF Rifa peg124.
Another curious thing was to replace the power resistors with the Manganin PBH 3/10W units from Issabellenhutte in Germany - not exactly the cheapest things but a much softer, more detailed sound than, for example the Caddocks or the Isotan power resistors.
I had quite a bit of fun with this remarkable amp - A bit of an adventure if you want something a bit different to the standard F3.
... my 2 cents
... ooahah, I thought as much.
Are there any best practices out there to match these IRFP240 for F3?
I bougth 2 sets of F3 boards from a german guy and 1 set of the LUIS where included.
I've already matched 2SK170 for my aleph ono and there was a wide spreading between the JFETs. I'ver odered 10 of the IRFP240 in hope they are much closer - is there a chance to get a proper triple?
The LED in the schematics: In my german translation is the LED marked as BLUE - I hope also a green one will work, with blue LEDs blood runs out of my eyes 😉
thanks & cheers
Hello, i am building my F3 amplifier, but i didnt see any schematic from the power supply
Anybody can help me out?
grtz
Anybody can help me out?
grtz
Hi,
just have look at Peter Daniels F3 PSU boards here http://www.diyaudio.com/forums/pass-labs/136037-f3-ps-options-using-pd-boards.html
it is quit simple: AC in - rectifier diodes - 15,000µF - R (0R5) - 15,000µF - DC unreg out.
These are boards for standard PSUs, you also can install a little LM 317 reg for preamps or s. th. like that, not nesseccary for F3.
I've also a set of these boards but there is not too much space for real big elctrolytics. I bought some Panasonic TS HA with 35mm diameter, the biggest I've found are Evox Rifa 22,000µF, but they won't fit into my case.
cheers
Thorsten
just have look at Peter Daniels F3 PSU boards here http://www.diyaudio.com/forums/pass-labs/136037-f3-ps-options-using-pd-boards.html
it is quit simple: AC in - rectifier diodes - 15,000µF - R (0R5) - 15,000µF - DC unreg out.
These are boards for standard PSUs, you also can install a little LM 317 reg for preamps or s. th. like that, not nesseccary for F3.
I've also a set of these boards but there is not too much space for real big elctrolytics. I bought some Panasonic TS HA with 35mm diameter, the biggest I've found are Evox Rifa 22,000µF, but they won't fit into my case.
cheers
Thorsten
You should also use caps with a higher voltage rating. For example 63 V to have some headroom. I think the original design uses 25 V caps but two caps are always connected in series to get 50 V rating.
Has anyone tried the paralleled silmics and decided to go with the original panasonic output cap afterwards? I went for the silmics right from the start but in the end felt I didn't quite get the right sound out from the amp even though everything else seemed to be alright (voltages etc. can't measure the output distortion). Would it be worth trying out the original combination of 15000uF Panasonic + 220uF silmic and 1uF PP?
Hi thanx for the info, but actually i am talking about another way off building the supply board.
First convert ac into dc with sckotty diodes then 22.000 uF - coil - 22.000 uF.
Anybody knows the value off the coil between the capacitors?
I like coil there better then resistors.
First convert ac into dc with sckotty diodes then 22.000 uF - coil - 22.000 uF.
Anybody knows the value off the coil between the capacitors?
I like coil there better then resistors.
Using a coil you will get better ripple filtering, but coils have some disadvantages in sound, because they are _slow_ in delivering current.
The value of the coil or resistor depends on the transformer, caps ESR you like to use, load at DC out and in case of the coil also the DC resistnace of the coil
There are some useful tools out ther in the web, just google PSU designer or so
cheers
Thorsten
Matchin IRFP240
Hi Folks,
for my F3 I've now matched the MOS FETs like Nelsons manual: http://www.firstwatt.com/pdf/art_matching.pdf
First I measured with a 470R and got a spread for Ugs from 3,524 V to 3,568V
At first I thought the ETs are broken because Nelson wrote in his manual Ugs should be between 4V and 4,57 V, but a look in the spec said between 2V and 4 V
Then I measured with a 13R6 power Res each FET exactly 2 min clamped on a heatsink at 720mA
So I got the following
Q5 Q2 Q3
Left: 4,403 4,424 4,425
Right: 4,402 4,422 4,422
Ugs is pretty close in my opinion.
What do you think of my choice for the Qs?
thanks & cheers
Thorsten
Hi Folks,
for my F3 I've now matched the MOS FETs like Nelsons manual: http://www.firstwatt.com/pdf/art_matching.pdf
First I measured with a 470R and got a spread for Ugs from 3,524 V to 3,568V
At first I thought the ETs are broken because Nelson wrote in his manual Ugs should be between 4V and 4,57 V, but a look in the spec said between 2V and 4 V
Then I measured with a 13R6 power Res each FET exactly 2 min clamped on a heatsink at 720mA
So I got the following
Q5 Q2 Q3
Left: 4,403 4,424 4,425
Right: 4,402 4,422 4,422
Ugs is pretty close in my opinion.
What do you think of my choice for the Qs?
thanks & cheers
Thorsten
It must be that either your electrons are dizzy from going in circles, or maybe your inductors are lazy.
Either way your experience is vastly different using inductors in a PS circuit are contrary to my experience and hundreds of builder reports.
I was thinking of using my chassis of aleph-x and power supply for a F3 build.Problem is the ps is 2x35 v. Would this be ok or too much? Off hand I do not see where any part is working over its voltage limit.
Well then you are lucky because you do not have to use a large value electro cap. A nice film in oil cap of around 100uf might be all you need.
I am looking for some F3 boards but no luck yet. Where did you get yours?
I have plenty of matched irf 240's.
Hi,
your PS has 2 X 35 V DC? That wouldn't be enough or you have a loss of output power.
A transformer with 2 X 35V AC would be OK if it has really two separate windings. You can use 1 winding for each channel. I had trouble with a 37-0-37V center tapped. Each chanel worked fine, but in stereo the two grounds stacked together and doubles the PSU voltage from 46V to 92V - C7 jumped right into my face
Stacking the two windings together you will get more than 100 V DC. A bias arround 47 V and a power consumption of arround 400 W - You'll have to buy a lot of alloy and real big caps!
I've bought mine 2nd hand but not used here in germany.
A nice solution for the output coupling is from here the next 20 posts.
My F3 is running since yesterday. I've builded as discussed there, but not testet yet. Currently I'm comparing F3 with Naim Audio NAP 250
If you are interessted in some details of my F3 - feel free
cheers
Last edited:
DC. That would make it 70V in series. A bit high but that would make about 150 W dissipation more or less.
I think my heatsinks should be enough as they are monoblocks with about .33C/W per side.The other consideration is if there is more distortion at that voltage. Perhaps it is too high for the LU.
I did not get where you got your pcb's. Are they still available?
Or maybe you could part with the extra set you have😀
I think my heatsinks should be enough as they are monoblocks with about .33C/W per side.The other consideration is if there is more distortion at that voltage. Perhaps it is too high for the LU.
I did not get where you got your pcb's. Are they still available?
Or maybe you could part with the extra set you have😀